24. Ecosystem Climate Manipulations
and chemical soil properties will adapt more
slowly. The carbon content of soil along a gradient
may vary because of differences in the amount of
time that has elapsed since soil formation began,
differences in vegetation communities with differing litter quality, and differing microbial activity
due directly to temperature and moisture. In one
study, soil mesofaunal abundance and diversity
showed dramatically different responses to manipulated soil microclimate, to interannual soil microclimate variation, and to spatial variation in soil
microclimate along a gradient (Harte et al. 1995b).
Chapin et al. (1995) found that longer-term responses (9 years) in a passive greenhouse experiment showed closer correspondence to patterns of
vegetation distribution along environmental gradients than did shorter-term (3 year) responses. Further study of ecosystem responses to manipulated
and natural climate variation can help to identify
when and how gradient analyses can substitute for
manipulation experiments (Pacala and Hurtt 1993;
Vitousek and Matson 1991).
Reciprocal transplants and common garden designs (e.g., Anderson et al. 1996) permit study of
selected climate factors, and can augment surrogate
gradient studies. For example, a reciprocal soil
transplant would interchange soil from a cool spot
and a warm spot on the gradient, thus reducing the
confounding influences of physical and chemical
differences in soil on response to climate change.
Climate influence on vegetation communities can
also be explored. Jensen (1993) used a common
garden design to estimate the relative effects of
source-site climate and transplant-site climate on
white fir (Abies concolor) sapling growth rates. The
transplant scheme has its limitations, not the least
of which is the intrinsically small scale. Reciprocal
transplants do not provide a whole-ecosystem perspective, but rather, may be used in conjunction
with techniques which give information at larger
scales.
Field Manipulations:
Warming Experiments
The object of most manipulation experiments is to
alter one or more climate parameters, such as air or
soil temperature, precipitation, radiation, or timing
of snowmelt. Changing anyone of these, however,
can affect some or all of the others as well, and
357
such secondary meteorological responses are not
always intended or predictable. Parameters of direct
concern to ecologists, such as soil moisture, are
likely to be jointly influenced by all of these secondary responses to the directly manipulated parameter. In this section, we discuss experimental
approaches for direct warming of ecosystems, and
discuss secondary responses as well.
Greenhouses
The simplest and most prevalent method of increasing temperature is a passive greenhouse. Passive
greenhouses have been deployed across a variety
of habitats, primarily at high latitude or high elevation, including Arctic tussock tundra, Antarctic
fellfield, subalpine meadows, Tibetan plateau, and
temperate grassland. Greenhouses are typically
used to elevate temperature, although they have
also been used to manipulate other climate variables, such as water (soil moisture and humidity),
light (level and spectral quality), and CO 2 • Greenhouses generally achieve a temperature increase of
about 2 to 6°C above ambient, depending on the
design and objectives of the study. In general,
greenhouses are inexpensive (which facilitates replication), simple to build, and low maintenance.
They can be used in remote areas without electrical
power, and in long-term simulations. Chapin et al.
(1995), for example, ran a greenhouse manipUlation
study for 9 growing seasons in an Arctic tundra
community. Greenhouses are seldom left in place
during the winter months, however, as they have
typically been deployed in harsh climates and cannot withstand winter conditions.
Greenhouses come in a variety of sizes and designs, from a variety of open-topped designs, to
horticultural cloches, to dome tents, to screens, to
classic greenhouses with glass, plastic, or fabric
panels (Fig. 24.1). An appropriate control might be
an open plot shaded with the same material (Debevec and Maclean 1993). The choice of materials
is as important as structural design; Debevec and
Maclean (1993) found significant differences in
temperature elevation, light attenuation, and gas
transmission among different combination of
greenhouse materials. Kennedy (1995a) lists several recent studies and notes their greenhouse design and the temperature elevation achieved. The
International Tundra Experiment (ITEX) (e.g.
and chemical soil properties will adapt more
slowly. The carbon content of soil along a gradient
may vary because of differences in the amount of
time that has elapsed since soil formation began,
differences in vegetation communities with differing litter quality, and differing microbial activity
due directly to temperature and moisture. In one
study, soil mesofaunal abundance and diversity
showed dramatically different responses to manipulated soil microclimate, to interannual soil microclimate variation, and to spatial variation in soil
microclimate along a gradient (Harte et al. 1995b).
Chapin et al. (1995) found that longer-term responses (9 years) in a passive greenhouse experiment showed closer correspondence to patterns of
vegetation distribution along environmental gradients than did shorter-term (3 year) responses. Further study of ecosystem responses to manipulated
and natural climate variation can help to identify
when and how gradient analyses can substitute for
manipulation experiments (Pacala and Hurtt 1993;
Vitousek and Matson 1991).
Reciprocal transplants and common garden designs (e.g., Anderson et al. 1996) permit study of
selected climate factors, and can augment surrogate
gradient studies. For example, a reciprocal soil
transplant would interchange soil from a cool spot
and a warm spot on the gradient, thus reducing the
confounding influences of physical and chemical
differences in soil on response to climate change.
Climate influence on vegetation communities can
also be explored. Jensen (1993) used a common
garden design to estimate the relative effects of
source-site climate and transplant-site climate on
white fir (Abies concolor) sapling growth rates. The
transplant scheme has its limitations, not the least
of which is the intrinsically small scale. Reciprocal
transplants do not provide a whole-ecosystem perspective, but rather, may be used in conjunction
with techniques which give information at larger
scales.
Field Manipulations:
Warming Experiments
The object of most manipulation experiments is to
alter one or more climate parameters, such as air or
soil temperature, precipitation, radiation, or timing
of snowmelt. Changing anyone of these, however,
can affect some or all of the others as well, and
357
such secondary meteorological responses are not
always intended or predictable. Parameters of direct
concern to ecologists, such as soil moisture, are
likely to be jointly influenced by all of these secondary responses to the directly manipulated parameter. In this section, we discuss experimental
approaches for direct warming of ecosystems, and
discuss secondary responses as well.
Greenhouses
The simplest and most prevalent method of increasing temperature is a passive greenhouse. Passive
greenhouses have been deployed across a variety
of habitats, primarily at high latitude or high elevation, including Arctic tussock tundra, Antarctic
fellfield, subalpine meadows, Tibetan plateau, and
temperate grassland. Greenhouses are typically
used to elevate temperature, although they have
also been used to manipulate other climate variables, such as water (soil moisture and humidity),
light (level and spectral quality), and CO 2 • Greenhouses generally achieve a temperature increase of
about 2 to 6°C above ambient, depending on the
design and objectives of the study. In general,
greenhouses are inexpensive (which facilitates replication), simple to build, and low maintenance.
They can be used in remote areas without electrical
power, and in long-term simulations. Chapin et al.
(1995), for example, ran a greenhouse manipUlation
study for 9 growing seasons in an Arctic tundra
community. Greenhouses are seldom left in place
during the winter months, however, as they have
typically been deployed in harsh climates and cannot withstand winter conditions.
Greenhouses come in a variety of sizes and designs, from a variety of open-topped designs, to
horticultural cloches, to dome tents, to screens, to
classic greenhouses with glass, plastic, or fabric
panels (Fig. 24.1). An appropriate control might be
an open plot shaded with the same material (Debevec and Maclean 1993). The choice of materials
is as important as structural design; Debevec and
Maclean (1993) found significant differences in
temperature elevation, light attenuation, and gas
transmission among different combination of
greenhouse materials. Kennedy (1995a) lists several recent studies and notes their greenhouse design and the temperature elevation achieved. The
International Tundra Experiment (ITEX) (e.g.